Soil Science Society of America Journal 65:1302-1306 (2001)
© 2001 Soil Science Society of America
DIVISION S-8—NUTRIENT MANAGEMENT & SOIL & PLANT ANALYSIS
Design and Ammonia-Recovery Evaluation of a Wind Speed-Sensitive Chamber System
M. L. Cabrera*,a,
D. E. Kissela,
R. C. Davisb,
N. P. Qafokua and
W. I. Segarsa
a Crop & Soil Sciences, Univ. of Georgia, Athens, GA 30602
b Instrument Shop, Univ. of Georgia, Athens, GA 30602
* Corresponding author (mcabrera{at}arches.uga.edu)
 |
ABSTRACT
|
|---|
A previously developed system to measure NH3 volatilization in the field consists of a vacuum pump, volatilization chambers, and acid traps to capture NH3. The vacuum pump of the system draws air through each chamber at a constant rate, which may be faster or slower than the wind speed outside the chamber. Because wind speed affects NH3 volatilization, it would be desirable to draw air through the chambers at a wind speed similar to that outside the chambers. The objective of this work was to improve the existing design by adding the capability of adjusting the speed at which air is drawn through each chamber. A hot needle anemometer was placed next to each chamber to measure wind speed at 1 cm above the soil surface. Wind speed was used to adjust an electronic valve so that the rate at which air was drawn through the chamber would match the external wind speed. Three N recovery studies were conducted with a sand–CaCO3 mixture in the field to determine if the system could be used to obtain quantitative estimates of NH3 volatilized. The NH3 volatilized in each study was estimated from (i) the loss of soil inorganic N and (ii) the N captured in the acid traps. In all three studies, the NH3 volatilized estimated from the acid traps (10–15% of applied N) was not different from the NH3 volatilized estimated from the loss of soil inorganic N. These results indicate that the system can be used to obtain quantitative measurements of NH3 volatilized under the conditions of these studies.
|
INTRODUCTION
|
|---|
SURFACE-APPLIED UREA FERTILIZERS commonly result in N losses through NH3 volatilization (McInnes et al., 1986; Al-Kanani et al., 1991; Nathan and Malzer, 1994). Understanding the magnitude of these losses under different conditions is important for proper fertilizer management. Several methods have been developed to measure NH3 volatilization in the field, with approaches ranging from micrometeorological techniques to chambers and wind tunnels (Harper, 1988). Micrometeorological techniques are typically considered the most accurate (Ferguson et al., 1988), but because of the large plots required they are not practical for simultaneous evaluation of several treatments. Chambers and wind tunnels require smaller plot sizes and thus can be used for studies with many treatments. One of the concerns with chambers and wind tunnels, however, is the potential alteration of environmental conditions, which in turn may affect NH3 volatilization. In an effort to minimize changes in environmental conditions, Kissel et al. (1977) developed an automated chamber system that periodically places a lid over an enclosed microplot while sweeping air through the chamber to sample the NH3 being volatilized. After sampling, the system opens the lid to expose the microplot to existing environmental conditions. Although this system minimizes environmental changes, a potential drawback is that air is drawn through the chamber at a constant speed, which may be faster or slower than the wind speed outside the chamber. Air drawn at faster speeds can increase NH3 volatilization, whereas air drawn at slower speeds can decrease NH3 volatilization.
The effect of wind speed on NH3 volatilization is clearly demonstrated by the data of Fillery et al. (1984), who found that the rate of NH3 loss from a flooded rice (Oryza sativa L.) field increased linearly with wind speed over the range of 0 to 8 m s-1. Ferguson et al. (1988) compared the chamber system of Kissel et al. (1977) to a micrometeorological method in a field that received surface-applied urea over wheat (Triticum aestivum L.) stubble. They found that the NH3 losses by the micrometeorological method were 7- to 16-fold larger than those by the chamber method. They attributed the lower losses observed with the chamber method to its lack of sensitivity to high wind speeds observed immediately after a rain. In contrast, Hargrove et al. (1987) found that a chamber system with constant air withdrawal rate overestimated the amount of NH3 volatilized from urea applied to a mulched soil when compared with 15N recovery. They attributed the overestimation of the chamber system to the existence of external wind speeds that were lower than that generated with their system (0.06 m s-1). This effect of wind speed may be particularly important in areas where low wind speeds are common and in situations where the presence of mulch or vegetation (such as in a forest) may reduce wind speeds near the soil surface. Under those conditions, it may be desirable to have a system that can adjust the speed of air withdrawal based on the external wind speed. Therefore, the objective of this work was to improve the system designed by Kissel et al. (1977) by adding the capability of matching air speed through the lid to the external wind speed.
|
MATERIALS AND METHODS
|
|---|
System Design
Our overall system design uses square chambers instead of the round ones used by Kissel et al. (1977). Each chamber is a stainless steel box (18 by 18 by 10 cm deep) with open ends that is pushed into the ground until the top of the chamber is flush with the soil surface (Fig. 1). A reversible electric motor (Model 6ZL32, Hurst Mfg., Princeton IN) mounted on the outside of the box opens and closes an acrylic lid at intervals regulated by a programmable controller (Toshiba T1-MDR40S, Toshiba International Corp., Houston, TX). When the lid is closed, a blower pump (Model R21102, Gast Mfg. Corp., Benton Harbor, MI) draws air through the chamber headspace (518 mL) into 650 mL of 0.05 M H2SO4 to trap the swept NH3. The air flow rate through the chamber headspace is determined by the vacuum generated by the pump, which in turn is controlled by an electronic valve (Model PC2000, Texsteam Inc., Houston, TX). A needle anemometer (Bland et al., 1995; modified 229 Probe, Campbell Scientific Inc., Logan, UT) located next to each chamber measures wind speed at a height of 1 cm above the soil surface, and a Campbell Scientific 21X datalogger adjusts the electronic valve to generate an air speed inside the chamber that matches the external wind speed. The process of adjusting the electronic valve involves three steps (Fig. 2). In Step 1, the 21X datalogger calculates the average external wind speed measured with the needle anemometer during the previous 5 min. In Step 2, the 21X datalogger outputs a proportional voltage to one of its continuous analog output channels, based on the average external wind speed. In Step 3, the voltage generated by the 21X is fed to the electronic valve to adjust the wind speed inside the chamber headspace (internal wind speed).
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 2. Steps in electronic valve adjustment. In Step 1, the datalogger calculates the average external wind speed based on anemometer readings. In Step 2, the datalogger outputs a proportional voltage based on the 5-min average external wind speed. In Step 3, the voltage output from the datalogger is fed to the electronic valve to adjust the wind speed inside the chamber head space.
|
|
Our design prototype consists of three chambers. Vacuum from the pump is directed to one chamber at a time by opening and closing solenoid valves (Type 18-33 A, Magnatrol Valve Corp, Hawthorne, NJ) placed between the pump and each of the acid traps. Consequently, the chambers are sampled one after the other. We used this design to minimize pump size requirements because simultaneous sampling of the three chambers would have increased the size of the pump required by threefold.
Every 15 min, the programmable controller activates the chamber motor to close the lid and sample the air for 115 s. While the lid is closed, the pump draws air from the chamber headspace and bubbles it through the acid trap. After the 115-s sampling, the controller reverses the chamber motor to open the lid, exposing the chamber to natural wind conditions. After each chamber is sampled, background air is bubbled through a control acid trap for 38.3 s to sample background NH3 in the air. Since there are three chambers in the system and background is sampled for 38.3 s after each chamber, background NH3 is sampled for a total of 115 s every 15-min interval, as is the case for each of the chambers. We used a sampling time of 115 s to minimize the effect of the lid on soil surface temperature. Longer periods could lead to an increase in soil temperature. We sampled every 15 min to capture changes in the rate of volatilization caused by diurnal changes in temperature.
Lid Design
The acrylic lid has a rectangular section joining a triangular section (Fig. 3). The rectangular section (19.5 cm wide by 25 cm long by 1.3 cm deep) covers the fertilizer-treated area when the lid is closed. The triangular section (34.5 cm long by 19.5 cm wide at the base by 1.3 cm deep) is closed at its top and bottom ends to create a conduit through which air is drawn from the treated area. At the beginning of the conduit there is a slit, which allows vacuum to be applied evenly to one side of the treated area. This is needed to obtain an homogeneous sweep of the treated area.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 3. Side and bottom view of the acrylic lid. The slot labeled "open" on the left side of the bottom view allows outside air in. The square area labeled "open" in the center of the bottom view closes over the treated area so that the air drawn in through the slot can sweep the NH3 volatilized from the treated area. The slit on the right side of the lid (side view) generates a uniform sweep of the treated area.
|
|
Response of Ammonia Volatilization to Wind Speed
To determine the response of NH3 volatilization to wind speed through the lid, we conducted a laboratory study with soil collected from an area mapped as Cecil sandy loam (clayey, kaolinitic, thermic Typic Kanhapludult). The soil was amended with 100 g CaCO3 kg-1 to increase its pH from 5.9 to 7.8 and to favor NH3 volatilization. After the soil was wetted to 0.10 g g-1, 4.9 kg of dry-weight equivalent soil was packed to a density of 1.5 Mg m-3 into an acrylic box (18.1 by 18.1 by 10 cm deep). Before initiating the experiment, the top 0.3 cm of soil was removed from the box, mixed with 1.17 g NH4NO3 (62.5 kg NH4–N ha-1) and packed back. Immediately after NH4NO3 was added, the acrylic lid was placed over the box and a vacuum pump was connected to the lid for 96 h. Air drawn through the lid was passed through an acid trap (150 mL 0.05 M H2SO4 changed every 12 h) to capture the NH3 volatilized, and was controlled by a flow meter (Dwyer Instruments Inc., Michigan City, IN) to generate different wind speeds. Separate studies were conducted with wind speeds of 0.003, 0.006, 0.012, 0.024, 0.042, 0.066, 0.096, and 0.135 m s-1. The whole system was set up in a growth chamber (23°C) so that the relative humidity could be maintained at 98% to prevent large losses of water. The results showed an increase in NH3 volatilization with an increase in wind speed, and a leveling off at 0.135 m s-1 (Fig. 4). The NH3 lost at the largest wind speed corresponded to 40% of the NH4–N applied. On the basis of these results, we designed our field system to generate a maximum wind speed of 0.25 m s-1 to allow for a margin of error in case the system responded differently to wind speed once deployed in the field. When the external wind speed is >0.25 m s-1, the field system generates a wind speed of 0.25 m s-1. The minimum wind speed that the field system can generate is 0.05 m s-1.
System Test
Three field studies, with three replications each, were conducted to determine if the system could be used to obtain quantitative recoveries of NH3 volatilized. The system was installed on the floor of a pine (Pinus taeda L.) forest located at the Whitehall Forest, University of Georgia (N 33° 53.500', W 83° 21.363'). Instead of soil, sand (Play Sand, Quikrete Companies, Atlanta, GA; 4% 1000–2000 µm; 53% 500–1000 µm; 33% 350–500 µm; 5% 250–350 µm, and 5% 75–250 µm) mixed with CaCO3 (100 g kg-1) and wetted to 0.1 g g-1 was used to obtain conditions that favor NH3 volatilization (pH = 7.8) and prevent N mineralization, immobilization, nitrification, and denitrification. A preliminary test confirmed that the sand mixture did not mineralize N or immobilize fertilizer N, and did not have any nitrification activity. A stainless steel tray (18 by 18 by 3.5 cm) was placed inside each of the chambers and 1.5 kg of dry-weight equivalent sand–CaCO3 mixture was packed into each tray. Subsequently, a 10-mL syringe (with needle) was used to apply 5 mL of an NH4Cl solution (324 mg NH4–N) to the surface of the sand–CaCO3 mixture. To determine the amount of N applied that was extractable with 1 M KCl, one extra chamber was prepared, treated, and extracted immediately after application. The amounts recovered immediately after application were 259 mg N in the first study, 257 mg N in the second study, and 283 mg N in the third study (78–86 kg N ha-1). The system was started immediately after N application and NH3 volatilized was collected in traps containing 650 mL of 0.05 M H2SO4. The first study was conducted at the end of September 1999 and lasted 19 h. The second and third studies were conducted in November 1999 and lasted 23 and 71 h, respectively. At the end of each study, the sand mixture from each chamber was extracted with 7 L of 1 M KCl for 30 min, and a sample of the supernatant volume was analyzed for inorganic N. Ammonia-N in the traps was also determined. The concentration of (NO2–N + NO3–N) in KCl extracts was determined by the Griess-Ilosvay technique after reduction of NO3 to NO2 with a Cd column (Keeney and Nelson, 1982). The concentration of NH4–N in traps and KCl extracts was determined by the salicylate-hypochlorite method (Crooke and Simpson, 1971). The amount of NH3 volatilized was estimated using two approaches: (i) from the difference between the inorganic N applied and the inorganic N remaining in the sand mixture (loss of inorganic N), and (ii) from the amount of N collected in the traps multiplied by 7.826. This multiplication factor was obtained by dividing the interval between samplings (15 min = 900 s) by the chamber sampling time (115 s). The amounts of NH3 volatilized were divided by the amounts of N applied to estimate the proportion of N volatilized. An analysis of variance was conducted and Fisher's LSD was used to compare the percentage of N volatilized estimated from the loss of inorganic N with the percentage of N volatilized estimated from N collected in the traps (SAS Institute, 1994).
|
RESULTS AND DISCUSSION
|
|---|
During Study 1, the wind speed 1 cm above the soil surface varied between 0.07 and 0.47 m s-1 and remained above the threshold of 0.135 m s-1 approximately 88% of the time (data not shown). The proportion of the applied N lost through NH3 volatilization was 11.7% when estimated from the loss of inorganic N and 12.1% when estimated from the N captured in the traps (Table 1). Similar to Study 1, wind speed during Study 2 showed a range of 0.08 to 0.47 m s-1 and remained above 0.135 m s-1 approximately 91% of the time. The percentage of N lost through NH3 volatilization was 10.5% when estimated from the loss of inorganic N and 10.1% when estimated from the N captured in the traps (Table 1). During Study 3, the wind speed varied between 0.06 and 0.47 m s-1 and stayed above 0.135 m s-1 about 87% of the time. The percentage of N lost as NH3 was 15.1% when estimated from the loss of inorganic N and 15.0% when estimated from the N captured in the traps (Table 1).
Study 3 had a larger NH3 loss than the other two studies because it had a duration of 71 h as opposed to 19 h for Study 1 and 23 h for Study 2. Nevertheless, all three studies had relatively high NH3 losses considering their short duration. We wanted to test the system under high rates of NH3 loss because we reasoned that if the system could perform well under high rates of loss, it would be expected to perform well under low rates of loss. This is so because the main challenge with systems that use acid traps to capture NH3 is to obtain complete NH3 scrubbing when the air drawn through a trap has a high NH3 concentration.
In all three studies, the amount of NH3 volatilized estimated from the N captured in the traps was not significantly different from that estimated from the loss of inorganic N (Table 1). Furthermore, when we added the inorganic N remaining in the soil to the NH3–N lost estimated from the acid traps, we obtained recoveries ranging from 98 to 102% of the N applied (data not shown). These results indicated that the method can be used to obtain quantitative recoveries of NH3 volatilized under the conditions of our study. It should be kept in mind, however, that our system used an artificial soil (sand and CaCO3) to facilitate NH3 volatilization and prevent other N transformations that could add or remove N from the inorganic pool. It is because of these conditions that we were able to use a mass balance of the applied inorganic N to estimate the amount of NH3 volatilized from the system. A mass balance of the applied inorganic N may not be suitable for a natural soil because other N transformations (i.e., mineralization, immobilization, denitrification), in addition to NH3 volatilization, may add or remove N from the inorganic pool. Nevertheless, if those N transformations that can affect the results are known to be negligible, the mass balance approach may be useful to estimate NH3 volatilized under natural conditions. For example, Hargrove et al. (1987) measured NH3 volatilized from bare and mulched soil by 15N recovery and by mass balance of the applied urea. For the bare soil, the NH3 loss expressed as a percentage of the urea-N applied was 18% by 15N recovery and 14% by mass balance. For the mulched soil, the NH3 loss was 41% by 15N recovery and 46% by mass balance. These results clearly indicate that for the conditions of their study, the mass balance approach was comparable to the 15N recovery method.
Wind speeds were above the maximum wind speed that can be generated by the system (0.25 m s-1) during 9% of the time in Study 1, during 36% of the time in Study 2, and during 23% of the time in Study 3. Because in all cases the system generated good recoveries of the amounts of NH3 volatilized, these results confirm the existence of a wind speed threshold under field conditions, which may be the same as that measured under laboratory conditions (0.135 m s-1). If that is the case, the capability of adjusting wind speeds was necessary for approximately 10% of the time in all three studies. It should be noted, however, that wind speeds during the summer are usually lower at this site. For example, wind speed measured during 10 d at the beginning of September 1999 was below 0.135 m s-1 approximately 49% of the time (data not shown). Under those conditions, the capability of adjusting wind speed may be even more important than in the present studies.
|
SUMMARY AND CONCLUSIONS
|
|---|
We developed a modified chamber system to measure NH3 volatilization in which the wind speed inside the chambers was matched to the external wind speed. Three recovery studies with a sand–CaCO3 mixture showed that the amount of N volatilized estimated with the system was not different from that estimated from the loss of soil inorganic N. These results indicated that the system can be used to obtain quantitative recoveries of NH3 volatilized under the conditions of these studies.
|
ACKNOWLEDGMENTS
|
|---|
We thank the Georgia Traditional Industries Program, Gold Kist and Southern State Cooperatives, Champion International, Weyerhaeuser, Bowater, Georgia Pacific, Westvaco, and U.S. Borax for their financial support. We are grateful to Dr. Larry Morris for arranging access to a pine forest at the Whitehall Forest of the University of Georgia, and to John Rema, Jane Raikes, and John Senter for excellent technical support.
Received for publication March 13, 2000.
|
REFERENCES
|
|---|
- Al-Kanani, G., A.F. MacKenzie, and N.N. Barthakur. 1991. Soil water and ammonia volatilization relationships with surface-applied nitrogen fertilizer solutions. Soil Sci. Soc. Am. J. 55:1761–1766.[Abstract/Free Full Text]
- Bland, W.L., J.M. Norman, G.S. Campbell, C. Calissendorf, and E.E. Miller. 1995. A transiently heated needle anemometer. Agric. For. Meteorol. 74:227–235.
- Crooke, W.M., and W.E. Simpson. 1971. Determination of ammonium Kjeldahl digest of crops by an automated procedure. J. Sci. Food Agric. 22:9–10.
- Ferguson, R.B., K.J. McInnes, D.E. Kissel, and E.T. Kanemasu. 1988. A comparison of methods of estimating ammonia volatilization in the field. Fert. Res. 15:55–69.
- Fillery, I.R.P., J.R. Simpson, and S.K. DeDatta. 1984. Influence of field environment and fertilizer management on ammonia loss from flooded rice. Soil Sci. Soc. Am. J. 48:914–920.[Abstract/Free Full Text]
- Hargrove, W.L., B.R. Bock, R.A. Raunikar, and W.J. Urban. 1987. Comparison of forced-draft technique to nitrogen-15 recovery for measuring ammonia volatilization under field conditions. Soil Sci. Soc. Am. J. 51:124–128.[Abstract/Free Full Text]
- Harper, LA. 1988. Comparison of methods to measure ammonia volatilization in the field. p. 93–110. In B.R. Bock and D.E. Kissel (ed.) Ammonia volatilization from urea fertilizers. National Fertilizer Development Center, Muscle Shoals, AL.
- Keeney, D.R., and D.W. Nelson. 1982. Nitrogen—Inorganic forms. p. 643–698. In A.L. Page et al. (ed.) Methods of soil analysis. Part 2. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
- Kissel, D.E., H.L. Brewer, and G.F. Arkin. 1977. Design and test of a field sampler for ammonia volatilization. Soil Sci. Soc. Am. J. 41:1133–1138.[Abstract/Free Full Text]
- McInnes, K.J., R.B. Ferguson, D.E. Kissel, and E.T. Kanemasu. 1986. Field measurements of ammonia loss from surface applications of urea solution to bare soil. Agron. J. 78:192–196.[Abstract/Free Full Text]
- Nathan, M.V., and G.L. Malzer. 1994. Dynamics of ammonia volatilization from turkey manure and urea applied to soil. Soil Sci. Soc. Am. J. 58:985–990.[Abstract/Free Full Text]
- SAS Institute. 1994. SAS/STAT user's guide. Version 6, 4th ed. SAS Inst., Cary, NC.
This article has been cited by other articles:
|
|
|
|
 
D. E. Kissel, M. L. Cabrera, N. Vaio, J. R. Craig, J. A. Rema, and L. A. Morris
Rainfall Timing and Ammonia Loss from Urea in a Loblolly Pine Plantation
Soil Sci. Soc. Am. J.,
September 1, 2004;
68(5):
1744 - 1750.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
